Aether Cosmology Research Group
The Metamorphosis
Paradox
A caterpillar seals itself in a case and is digested down to a broth. Something with a different body plan climbs out and lays eggs for a creature it no longer is. Adaptive evolution explains a trait by the advantage it gives the organism carrying it. Here the organism is destroyed before the trait does a single day of work.
The Problem
Worse than the chicken and the egg
The chicken and the egg is a puzzle about ordering and nothing more. A chicken and its egg are one body plan at two ages, so the regress collapses the moment you allow an ancestor that was almost a chicken. Metamorphosis has no such exit. The two stages are not two ages of one body plan. They are two different animals, and the first one has to be killed and taken apart before the second one can exist.
A caterpillar cannot evolve wings, because a caterpillar with wings is a worse caterpillar, and the caterpillar is the only thing selection can see.
The facts, stated without dispute
Nothing here is contested by anyone. It is textbook developmental biology, taught in every entomology course. The whole argument runs on these four facts and nothing else.
The cycle
Egg, then larva (caterpillar), then pupa, then imago (butterfly), then egg again. Four forms, one genome.
Histolysis
Inside the pupa the larval body is taken apart by programmed self-digestion. Muscle, gut and cuticle are reduced to a broth. The caterpillar is not remodelled. It is consumed.
Imaginal discs
The adult is built from cell clusters set aside in the embryo. They sit inert for the whole of larval life and do not one day of work for the caterpillar that feeds them.
The hormonal switch
Juvenile hormone holds the larval program shut. When it falls, ecdysone fires the pupal program (Broad-Complex) and then the adult program (E93). Wrong order, dead insect.
Two animals, one genome, run in a fixed sequence. From the moment it hatches the caterpillar carries the complete blueprint for a second animal, physically present inside it, switched off, earning nothing, eating its food.
The paradox stated formally
Selection acts on a trait only through the survival or reproduction of the organism carrying it. Walk the cycle and ask at each stage what selection can actually see. The answer at every stage is nothing.
What selection can see at each stage
The caterpillar does not reproduce, so every caterpillar trait is visible to selection only through a butterfly that does not yet exist. The imaginal discs do nothing for the caterpillar. Not a little. Nothing. They eat its food and return no work, so in the caterpillar's own economy they are a straight loss. The butterfly does not produce butterflies, it produces eggs that produce caterpillars. The egg cannot fertilise itself.
Every stage exists only to build the next one. Selection pays out at reproduction, and reproduction sits two whole animals downstream of the traits that have to be paid for.
The chicken escape does not exist here
One sentence solves the chicken problem: the egg came first, laid by something almost a chicken. Put that sentence against metamorphosis and it dies on contact.
| Chicken and egg | Caterpillar and butterfly | |
|---|---|---|
| Relationship of the two stages | One body plan at two ages | Two different animals |
| Anatomical distance | Continuous growth, no reorganisation | Different mouthparts, legs, gut, wings, locomotion, diet, habitat |
| Does the earlier stage survive? | Yes, it becomes the later stage | No. It is digested. |
| Can you name an almost-version? | Yes, a slightly less chicken-like bird | No. A partly demolished larva is a corpse. |
| Does a small change help? | Yes, incremental improvement is available | No. Nothing pays out until the entire system is finished. |
A caterpillar and a butterfly are further apart anatomically than a caterpillar and a worm. You cannot say an almost-butterfly laid it, because the intermediate the argument needs is not a slightly different butterfly. It is an animal that starts the demolition with nothing to rebuild from. That is not a transitional form. It is a body in a bag.
The half-metamorphosis problem
Gradualism demands two things of every step: it must be viable, and it must be an improvement. Name the fifty percent version of metamorphosis and then look at what you have named.
An animal that digests its own musculature and gut while holding only part of the machinery needed to rebuild them is not a poorly adapted organism competing at a disadvantage. It is dead inside the case, and it leaves nothing behind for selection to act on.
The demolition system and the reconstruction system have to arrive together and already matched to each other. So do the hormone timers that fire them in the right order. Fire the demolition early and the larva liquefies with no discs ready to use it. Fire the rebuild first and there is neither material nor space to build in. Either way the animal ends as a sealed bag of fluid.
Neither half is worth anything on its own, so neither half can be stockpiled while waiting for the other to turn up. The mechanism on offer, useful parts accumulated one at a time, cannot build a system whose parts are lethal in isolation.
And there is no second attempt. A failed pupation does not yield a slightly worse butterfly that limps on and breeds. It yields a dead insect in a sealed case. Every wrong ordering of the switches deletes the lineage that tried it, so the ordering cannot have been found by trial.
The core contradiction
Adaptive evolution explains structures by the advantage they give the organism that carries them. Metamorphosis requires structures that give their carrier no advantage, that stay dormant for its entire life, and that only ever function inside a different animal, after the carrier has been digested.
Nothing a caterpillar does can select for an imaginal disc. The disc is paid off only by the success of an animal the caterpillar does not survive to become.
That is not variation and selection. That is a blueprint filed in advance, kept and fed for a whole lifetime on the strength of what it will one day be used for. Selection has no such facility.
The identity problem
Push one step further and ask what the unit of selection even is here.
The caterpillar dies. Its body is unmade. What climbs out is built from a cell lineage that never took any part in being a caterpillar.
If the caterpillar is the organism, it leaves no descendants, because it is destroyed. If the butterfly is the organism, then most of its life was spent as a different animal sharing almost none of its anatomy. Calling the two one organism is a filing convention, not an observation. The definition is doing the work the evidence was supposed to do.
Caterpillars trained in the fifth instar keep the aversion as moths. Caterpillars trained in the third instar keep it through two molts and then lose it in the case, because the embryonic lobe that held it is pruned during pupation (Blackiston, Silva Casey and Weiss, 2008). The demolition is not indiscriminate. It selects which parts of the brain to delete and which to carry over, and the caterpillar built one of those parts only to have it destroyed.
Objections and answers
The four standard replies, and what each one leaves untouched.
Berlese's hypothesis: the larva is just an embryo that hatched early.
This relabels the problem. It does not explain how an embryonic stage acquired an independent feeding anatomy, a working gut, functional legs and a complete behavioural repertoire, none of which an embryo has or needs. It renames the larva and leaves the demolition and rebuild machinery exactly where it found it, which is the part that needed explaining.
Look at hemimetabolous insects, they show a gradual series.
A nymph is a small adult. It has the adult body plan, wing pads included, and it grows into itself. That is growth, not metamorphosis. There is no series of living forms running from nymph to pupa. The comparison assumes the transition it was brought in to demonstrate. That is not evidence, it is a change of subject.
It evolved because splitting the niche is advantageous. The larva eats leaves, the adult drinks nectar, so they do not compete.
That describes the payout once the system is already finished. It is not a path to it. Every step on the way is lethal or costly, and selection cannot spend in advance against a return two animals down the line. Naming a destination does not build the road.
The genes were already there. It is just regulatory reordering.
Then say where the ordering came from. A sequence of switches that must fire in a fixed order, with a lethal outcome for every wrong order, is exactly the kind of thing that cannot be tuned by trial across generations. Reordering is the claim, not the explanation.
The argument compressed
- Selection preserves traits by their effect on the organism that has them.
- Imaginal discs do nothing for the caterpillar that carries them except cost it food.
- Therefore imaginal discs cannot have been preserved by selection acting on caterpillars.
- The butterfly that would benefit cannot exist until the discs already exist and already work.
- Therefore the mechanism requires the result before it can produce the result.
- Every partial version of the system kills its carrier, so it cannot be assembled in steps.
- Therefore complete metamorphosis is not explicable by the accumulation of individually advantageous traits.
The caterpillar builds a tomb, is digested inside it, and something else climbs out holding a complete set of instructions for a body that no ancestor of that body ever wore. Then it lays eggs for a creature it no longer is. Call that selection if you like. It selected nothing.
References
Sources
- Berlese, A. (1913). Intorno alle metamorfosi degli insetti. Redia, 9, 121-136.
- Blackiston, D. J., Silva Casey, E., & Weiss, M. R. (2008). Retention of memory through metamorphosis: Can a moth remember what it learned as a caterpillar? PLoS ONE, 3(3), e1736.
- Truman, J. W., & Riddiford, L. M. (1999). The origins of insect metamorphosis. Nature, 401, 447-452.
- Ureña, E., Chafino, S., Manjón, C., Franch-Marro, X., & Martín, D. (2016). The occurrence of the holometabolous pupal stage requires the interaction between E93, Krüppel-Homolog 1 and Broad-Complex. PLoS Genetics, 12(5), e1006020.
- Williamson, D. I. (2009). Caterpillars evolved from onychophorans by hybridogenesis. Proceedings of the National Academy of Sciences, 106(47), 19901-19905.